Technical Notes Anal. Chem. 1994,66, 2175-2179
Microjet Electrode: A Hydrodynamic Ultramicroelectrode with High Mass-Transfer Rates Julie V. Macpherson, Scott Marcar, and Patrick R. Unwin’ Department of Chemistry, University of Warwick, Covenrry CV4 7AL, U.K.
The microjet electrode (MJE)-a new hydrodynamic ultramicroelectrode-is described. In the MJE, a jet of solution is fired at high velocities (up to 25 m s-I) through a fme nozzle positioned directly over a disk ultramicroelectrode (UME) with the aid of micropositioners. The electrode is shown to be characterized by well-defined, variable, and high mass-transfer rates under steady-statevoltammetric conditions (upto 2 orders of magnitude greater than the conventional diffusion rate) and thus has considerable promise for the study of fast electrode kinetics. A new hydrodynamic electrode, in which a jet of solution is fired at high velocities (mean values up to 25 m s-l) from a capillary with a radius between 30 and 60 pm at a disk ultramicroelectrode (UME), is described. The resulting “microjet electrode” (MJE) is shown to be characterized by well-defined, variable, and high mass-transfer rates under steady-statevoltammetric conditionsand thus has considerable promise for the study of fast electrode kinetics. The advent of UMEs-electrodes with at least one dimension in the micrometer or sub-micrometer r a n g e h a s had a major impact on the voltammetric investigation of electrode kinetics and coupled chemical reactions.’ In particular, the reduced capacitative and resistive effects associated with small electrodes have allowed large perturbation methods, such as cyclic voltammetry and potential step chronoamperometry,to be extended into the sub-microsecond time domain.2 This approach has found widespread applica(1) For reviews, see for example: (a) Wightman, R. M.; Wipf, D. 0. In Electroanalytical Chemistry; Bard, A. J., Ed.;Marcel Dekker: New York, 1989; Vol. 15, p 267. (b) Heinze, J. Angew. Chem., Int. Ed. Engl. 1993,32, 1268. (c) Montenegro, M. I., Queirb, M. A., Daachbach, J. L., Eds. Microelectrodes: Theory and Applications; NATO ASI, Series E., Vol. 197; Kluwer: Dordrecht, 1991. (2) (a) Howell, J.O.; Wightman,R.M.Anal. Chem. 1984,56,524. (b) Wehmeyer, K. R.; Wightman, R. M. Anal. Chem. 1985, 57, 1989. (c) Howell, J. 0.; Kuhr, W. G.; Ensman, R. E.; Wightman, R. M. J. ElectroanaI. Chem. Interfocial Electrochem. 1986.209.77. (d) Montenegro, M. I.; Pletcher, D. J. Electroanal. Chem. Interfacial Electrochem. 1986, 200, 371. (e) Fitch, A.; Evans, D. H. J. Electroanal. Chem. Interfacial Electrochem. 1988,202, 83. (0 Andrieux, C. P.;Garreau, D.; Hapiot, P.; Pinson, J.; SavCant, J. M. J. Electroanal. Chem. Interfacial Electrochem. 1988, 243, 321. (g) Wipf, D. 0.;Kristensen, E. W.; Deakin, M. R.; Wightman, R. M. A M / . Chem. 1988,60.306. (h) Wipf, D. 0.;Wightman, R.M. J. Phys. Chem. 1989,93, 4286. (i) Andrieux, C. P.; Garreau, D.; Hapiot, P.; SavCant, J. M. J . Electroanal. Chem. Interfacial Electrochem. 1988,248,447. 0) Andrieux, C. P.; Hapiot,P.;SavQnt, J. M. J.Phys. Chem. 1988,92,5987. (k) Andrieux, C. P.; Hapiot, P.; Saveant, J. M. J. Phys. Chem. 1988.92.5992. (I) Garreau, D.; Hapiot, P.; SavCant, J. M.J.Electroanal. Chem. Interfacial Electrochem.
0003-2700194/03652175804.50/0 0 1994 American Chemical Society
tion although, at very short times in cyclic voltammetry, thecontribution to the electrode response from capacitative currents can become significant, making background subtraction procedures mandat~ry.~ Alternatively, since the steady-state diffusion-limited current density at an UME depends on the reciprocal of the electrode radius (for a disk electrode)? high mass-transfer rates may be obtained by using small electrodes. This has opened up the possibility of kinetically characterizing rapid reactions under steady-state condition^,^*^ with the advantage that the current response is largely free from resistive and capacitative complications. There are, however, the following minor drawbacks: (i) a range of electrode sizes is needed to achieve variable mass-transfer rates and (ii) when electrodes are sub-micrometer in size (so-called %anodes” 6, their geometries, which are difficult to determine, may make them unsuitable for the analysisof rapid electron-transfer kinetic^.^ More recently, Bard and co-workers have demonstrated that both high and variable diffusion rates may be obtained at UMEs by using the positive feedback mode of the scanning electrochemical microscope (SECM)? operating as a variable gap thin layer cell. This approach, in which diffusion rates to an UME may be increased by 1 order of magnitude? has been successfully employed to study both rapid electrontransfer kinetics9and the kineticsof chemicalprocessescoupled
(3) (4) (5)
(6) (7)
(8) (9)
1990,281,73. (m)Andrieux,C.P.;Hapiot,P.;SavCant, J. M.ElectrwMlysis 1990,2, 183. (n) Andrieux, C. P.; Audebert, P.; Hapiot, P.; SavCant, J. M. J . Am. Chem. Soc. 1990, 112, 2439. ( 0 ) Andrieux, C. P.; Audebert, P.; Hapiot, P.; SavCant, J. M. Synth. Met. 1991, 43, 2877. (p) Hseuh, C. C.; Brajter-Toth, A. Anal. Chem. 1993, 65, 1570. See, for example: Wightman, R. M.; Wipf, D. 0.Acc. Chem. Res. 1990,23, 64 and referencts therein. Saito, Y. Reo. Polarogr. 1968, IS, 177. (a) Bond, A. M.; Oldham, K. B.; Zoski, C. G. Anal. Chim. Acta 1989,216, 177. (b) Zhang, Y.; Baer, C. D.; Camaioni-Neto, C.; OBrien, P.; Sweigart, D. A. Inorg. Chem. 1991, 30, 1682. (c) Oldham, K. B.; Zoski, C. G. J. Electroanal. Chem. Interfacial Electrochem. 1988, 256, 11. Penner, R.M.; Heben, M. J.; Longin, T. L.; Lewis, N. S . Science 1990.250, 1118. (a) Baranski, A. S. J. Electroanal. Chem. Interfacial Electrochem. 1991, 307,287. (b) Oldham, K. B. Anal. Chem. 1992,64,646. (a) Bard, A. J.; Fan, F.-R. F.; Kwak, J.; Lev, 0.Anal. Chem. 1989,61,1794. (b) Bard, A. J.; Fan, F.-R. F.; Pierce, D. T.; Unwin, P. R.; Wipf, D. 0.;Zhou, F.Science 1991, 254, 68. (a) Wipf, D. 0.;Bard, A. J. J. Electrochem. Soc. 1991,138,469. (b) Bard, A. I.; Mirkin. M.V.; Unwin, P.R.; Wipf, D. 0. J , Phys. Chem. 1992, 96, 1861. (c) Mirkin, M. V.; Bard, A. J. J. Electrochem. Soc. 1993,139,3535. (d) Mirkin, M. V.; Bulb-, L.0. S.; Bard, A. J. J. Am. Chem. Soc. 1993, 115,201. (e) Mirkin, M. V.; Richards, T. C.; Bard, A. J. J. Phys. Chem. 1993, 97, 7672.
AnalyficalChemistry, Vol. 66, No. 13,July I, 1994 2175
s;tion
, fl4l-k
j
reference electrode
silicone rubber connector
1
capillary connected to x, y and z axis positioners
+solution outlet
plexiglass cell
(to waste)
Breadboard
Flgure 1. Schematic drawing (not to scale)of the microjet electrode.
to electron transfer.10 In this paper we suggest a new approach for achieving enhanced and variable steady-state mass-transfer rates at UMEs (up to almost 2 orders of magnitude greater than the conventional diffusion rate). This involves the use of a highvelocity microjet of solution which is fired through a small capillary positioned directly over the electrode with the aid of micropositioners. In the present study, experimental conditions are such that the diameter of the jet is around 3-5 times that of the disc electrode and the geometry thus representsa radical miniaturization of the conventional "walltube" electrode (WTE).' Under most practical hydrodynamic conditions, it is shown that mass transfer to the MJE is similar to that to the WTE, i.e., the rate of forced convection is sufficient for the diffusional edge effect at the UME to be suppressed and for the transport-limited current to be as predicted for a "uniformly accessible" WTE." Thus, in addition to the advantages of high and variable mass transport rates, the description of mass transport to the MJE is greatly simplified compared to the situation when mass transport is by diffusion alone. EXPERIMENTAL SECTION Apparatus and Instrumentation. A schematic diagram illustrating the essential features of the microjet electrode and cell is given in Figure 1. The cell, designed so that the base, main body, and lid were readily detachable, was constructed from plexiglass and had a volume of about 25 cm3. The UME, a 25-pm-diameter Pt (Goodfellow, Cambridge, U.K.) disk electrodewhich was fabricated and polished as described elsewhere,*J2was secured vertically in the base of the cell, via a holedrilled through the center, and positioned such that it sat below the level of the solution outlet. The nozzle of the MJE was constructed by drawing a borosilicate glass capillary (2.0-mm o.d., 1.16-mm i.d.) with the aid of a Narishighe (Japan) PB7 micropipette puller to produce a fine tip. The inner diameter of the tip of the capillary (typically in the range 60-120 pm) was measured directly with a precision of 1 pm using an optical microscope (Olympus (10) (a) Unwin, P. R.;Bard, A. J. J. Phys. Chem. 1991, 95,7814. (b) Zhou, F.; Unwin, P.R.; Bard, A. J. J. Phys. Chem. 1992, 96, 4917. (1 1) Chin, D. T.; Tsang, C. H. J. Electrochem.SOC.1978,125,1461 and references
therein. (12) Macpherson, J. V.; Unwin, P. R. J . Phys. Chem. 1994, 98, 1704.
2176 AnalyticalChemistry, Vol. 66, No. 13, July 1, 1994
BH2). The capillary was mounted on an x,y,z micropositioning stage which was used to control its position relative to the UME surface. The distance between the nozzle exit and the UME was generally established by initially placing the capillary in contact with the UME and then moving the capillary a known distance in the direction normal to the electrode using the micropositioning devices described below. The nozzle was centered above the electrode by scanning the capillary in the x-y plane, with the electrolyte flowing, until the transport-limited current attained its maximum value (see below). Two types of stages were employed in the MJE, in order to test the positioning requirements. Initially positioning was achieved with a closed-loop system, comprising Burleigh Instruments' (Fischer, NJ) TSE-75 integrated stages controlled by a 6200-3-3 controller interfaced to an IBMcompatible 486DX personal computer via a 660 interface card. This allowed reproducible positioning of the nozzle in the x, y, and z directions at a level of about 100 nm. In later experiments, a low-cost manual system comprising a Newport Corp. (Fountain Valley, CA) 46 1-xyz stage was employed. In this case the z axis was controlled by a differential micrometer (model DM-13) while fine adjustment screws (model AJS-1) were used to move the x- and y-axes. In the former case, stages were mounted on a Newport CSD series breadboard, which in turn was placed on a custom-built marble bench incorporating vibration isolators. In the latter case an aluminum breadboard supported on polystyrene foam was used to support the micropositioning stages. In the majority of experiments, flow of electrolyte through the nozzle of the MJE was achieved using a simple gravity feed system in which electrolyte was fed from a 500 cm3 reservoir, positioned at various heights above the outlet of the jet (typically 0.3-1.5 m), via several meters of PTFE tubing (15" bore). When slower flow rates were required, the flow system also contained a restricter (a glass tube of 0.2or 0.3-mm bore and about 30 cm in length). The dependence of the electrolyte volume flow rate on reservoir height was calibrated by measuring the mass of electrolyte solution flowing through the nozzle in a fixed period (and then converting this to the corresponding volume). Using the gravity-feed procedure, flow rates in the range to 5 X cm3 s-l were attainable for a typical 2X nozzle diameter of 100pm. In a few preliminary experiments, designed to examine the MJEvoltammetric response at higher solution velocities, flow was achieved with a Gilson (Villiersle-Bel, France) Minipuls 3 peristaltic pump. All measurements were made at 298 K using a two-electrode arrangement with a silver/silver chloride reference electrode. The UME potential was controlled with a purpose-built triangular wave generator (Colburn Electronics, Coventry, U.K.) and the current measured with a home-built current follower (gains of 10-5-10-9 A/V). Current-potential characteristics were recorded directly on an x-y recorder (PL3, Lloyd Instruments, Southampton, U.K.). Materials. Ferrocyanide solutions (at a concentration of mol dm-3) were prepared from potassium ca. 2 X ferrocyanide trihydrate (Aldrich, ACS grade) and Milli-Q reagent water (Millipore Corp.). Potassium chloride (Fluka,
1.2
30 25
I .-I
.o
20
1.1
4E
2 1.0
15
E. 10 0.9
0.6
0.0
1.5
3.0
4.5
6.0
1.5
H/d
Flguro 2. Effect of the lateralpoewOn of the nozzle, relathre to the center of the didc u k r m i c r w o d e , on thetransport-lknitedament at the mlcrojet electrode. The end of the nozzle was positroned 340 pm above the electrode surface, end a flow rate of 0.0181 cm9 s-' was employed.
Figure 3. Effect of the distance, H, between the nozzle exit and the
puriss p.a.) at a concentration of 0.5 mol dm-3 served as the background electrolyte.
impressive enhancements in current can be obtained with the impinging jet (even at moderate electrolyte flow rates) and that the required precision in lateral positioning, in this particular case, is of the order of the electrode diameter (25 rm) . The effect of the distance, H, between the nozzle exit and the electrode surface on the transport-limited current at the MJE was similarly investigated by carefully placing the nozzlein contact with the electrode surface, having centered it to the position of maximum current, and then measuring the transport-limited current as a function of electrodenozzle separation. The results of this exercise for a nozzle of diameter d = 1 10pm and flow rate of 0.0338 cm3s-I are given in Figure 3. The data have been normalized by dividing the microjet currents by the value measured at a distance H = d (corresponding in this case to I'MJEII'UME = 30.8), while the distance, H,has been normalized with respect to the nozzle diameter. Figure 3 demonstrates that, for most of thedistances considered, there is only a weak dependence of the transportlimited current on H / d . This observation is in agreement with the extensive studies of Chin and Tsang, with conventional-sized WTEs under laminar flow conditions, who found (empirically) an (H/d)-0.054 dependence of the transportlimitedcurrent for 0.2 < (H/d) < 6.0." For comparison with the experimental data, the behavior predicted by Chin and Tsang is given by the solid line in Figure 3. For experiments aimed at determining the effect of flow rate on the MJE transport-limited current, we were interested in elucidating to what extent mass transfer could be described by existing models for WTEs." For such electrodes, of conventionalsize, mass transfer, under laminar flow conditions, is described by expressions of the following general form:"
RESULTS AND DISCUSSION In order to establish the nature of mass transfer to the MJE, experiments were concerned with determining the effect of the position and flow rate of the impinging jet of solution on the transport-limited current at the UME for a simple electron-transfer process. For this purpose, the oxidation of hexacyanoferrate(I1) ions Fe(CN),'
- e-
-
Fe(CN),*
was studied at a 25-pm platinum MJE, with a range of nozzle diameters and flow rates. For measurements with both stationary and flowing solutions, the transport-limited current was determined by scanning the UME potential from around 0.00 to 0.70 V vs silver/silver chloride reference electrode (at a scan rate in the range 10-50 mV s-I) so as to record a full steady-state current-voltage curve. For MJE measurements, the nozzle was centered above the UME by first recording a voltammogram for hexacyanoferrate(I1) oxidation in stationary solution. With the electrode held at a potential corresponding to the transportlimited current, the jet was then turned on at the desired flow rate, and the nozzle was scanned in the x-y plane above the electrode until thecurrent attained its maximum value. Given the flow profile from the nozzle,11 this corresponds to the center of the jet-where the solution velocity is at its maximum-impinging on the center of the electrode. Thesensitivity of the electrodecurrent to the lateral position of the jet was readily determined by first centering the nozzle to the position of maximum transport-limited current and then measuring the current as a function of displacement in the x or y direction. Typical results, showing the jet current, I'MJE, normalized with respect to the diffusion-limited current in stationary solution, iUME, as a function of lateral displacement are shown in Figure 2. The data were obtained with the closed-loop micropositioning instrument using a nozzle of diameter d = 95 pm positioned 340 pm from the electrode surface and with an electrolyte flow rate of 0.0181 cm3 s-l (under gravity feed conditions). Figure 2 demonstrates that
electrodesvtaceonthetransporNMtedcvnmtat~MIE.Waswed currents, (O),have been normalizedwlth respectto that msur8d at a distance H = d, aH=d). Data relate to a nozzle of diameter d = 1 10 pm and a solution flow rate of 0.0338 cm9 s-l. The soHd llne represents the behavior predicted for a conventknecslzed walCtube electrode" (seetext for details).
Sh = ~tRe'/*Sc'/~flH/d) (1) In eq 1, a is a constant coefficient andf(H/d) is a distancedependent function; Sh, Re, and Sc are the Sherwood, Reynolds, and Schmidt numbers, given by Sh = k M d / D
(2)
Re = dU/u
(3)
35
1
90
1
30
0 1
0.00
0.05
0.10
0.15
0.20
0.25
(v,/cm3 Flgw 4. Analysis of transport-lknlted current-flow rate data for the MJE in terms of eq 7. Data relate to (0)d = 90 pm and H = 200 pm and (0)d = 118 pm and H = 300 pm.
SC= v / D (4) where kMT is the mass-transfer coefficient (cm s-l), D is the diffusion coefficient of the electroactivespecies, 8 i s the mean solution velocity (cm s-l), and Y is the kinematic viscosity of the solution. The transport-limited current for the uniformly accessible MJE is = ~MT~F(~u')c* (5) and the diffusion-limitedcurrent at an ultramicrodisk electrode is13 iMjE
iUME 4nFDac* (6) where, in eqs 5 and 6, Fis Faraday's constant, n is the number of electrons transferred per redox event, u is the electrode radius, and c* is the bulk concentration of the electroactive species. It follows from eqs 1-6 that, for the MJE, under conditions of uniform accessibility
and thus iMJE/i"ME should be proportional to the square root of the volume flow rate, Vf= r(d/2)%. Figure 4 shows data obtained from experiments with a 25-pm-diameter electrode and two nozzle diameters, d = 90 and 116 pm. In these cases, positioning was achieved using the closed-loop system, and the nozzle-electrode separations, H,were respectively 200 and 300 pm. It can be seen that, over the range of flow rates investigated, a good linear relationship holds between iMJE/ ~ U M Eand Vf1/2.The slopes of the lines yield values for af(H/ d)of 1.39 (d = 116-pm nozzle) and 1.41 (d = 90-pm nozzle), using values of D = 6.5 X I v cmz s-'-determined from eq 6 for voltammetric measurements in stationary solutions-and v = 0.0089 cm2 s-l. Both values of the product af(H/d) are in excellent agreement with the values a = 1.51 and f ( H / d ) = (H/d)-0.054 (8) determined by Chin and Tsang for the WTE." Given the lateral positioning requirements of the MJE (identified in Figure 2) and the weakdependenceof transportlimited current on the electrode-nozzle separation (Figure 3), we investigated whether the simple manual positioning 2178
AnelytlcelChemistry, Vd. 66, No. 13, July 1, 1994
0
I
I
I
I
I
100
200
300
400
500
v,l/2 d-l/2 (n/d)-O.084
(s-l/2)
FIguro 5. Anatysk of transportcurent-fkw rate data for the MIE in terms of eqs 7 and 8. Data relate to d = (0)84 pm, (V)96 pm, and (0)104 pm. In an cases the exlt of the nozzle was about 500 pm above the electrode surface.
system, described above, could give sufficient accuracy to be deployed in the MJE. Figure 5 is a composite plot, in terms of eqs 7 and 8, of the results obtained with this system using several nozzle diameters (104,96, and 64 pm) positioned ata distance of about 500 pm above a 25-pm-diameter electrode. For the 64-pm-diameter nozzle, flow was achieved using a peristaltic pump,13 while a gravity feed system was employed for the other two cases. The slope of the line yields a value of a = 1S O , in excellent agreement with that deduced by Chin and Tsang.11 It is clear from Figure 5 that enhancements of almost 2 orders of magnitude in the diffusion-limited current at an UME can be achieved through the addition of convection in the form of an impinging jet. It follows from eqs 5 and 6 that the highest mass-transfer rate in Figure 5 (I'MJEIIUME = 82.5) corresponds to a mass-transfer coefficient of about 0.55 cm s-I. An UME with a radius of about 150nm would be required to achieve the same mean rate of mass transfer in stationary solution (eq 6), while a rotating disc electrode would have to rotate at speeds of more than 200 OOO Hz in order to achieve the same mass-transfer rate.14 The latter is impossible to achieve practically. The highest mass-transfer rate for the data in Figure 5 was obtained with a volume flow rate of just 5.94 X 10-2 cm3 s-l and corresponds to a Reynolds number (eq 3) of 1.33 X 103. This is within the regime (Re < 2000) under which flow is 1aminar.l' Furthermore, since it should be possible to reduce the size of both the nozzle and electrode employed in the MJE, higher mass-transfer rates should be attainable under laminar flow conditions. However, it should be noted that mass transfer in the WTE geometry can readily be modeled under turbulent flow conditions" and thus Re < 2000 should not necessarily represent a limit in MJE studies. CONCLUSION The microjet electrode (MJE) represents a new approach for achievingsignificantlyenhancedand variable mass-transfer (13)
The peristaltic pump produced slight pulsations in the flow, reflected by an
oscillation in the transport-limited current of ca. 5% of the mean value. However,this did not introduceany problems in measuringthe mean transportlimited current. (14) Bard, A. J.; Faulkner, L. R. Elecrroehemicul Methods; Wiley: New York, 1980; p 283.
rates at UMEs (almost 2 orders of magnitude greater than the diffusion rate in stationary solutions). The electrode is characterized by well-defined mass transfer which, under most conditions, is similar to that for the conventionally-sizedwalltube electrode (WTE), Le. the electrode is “uniformly accessible”, which greatly simplifies the description of mass transfer. At present, the highest mass-transfer coefficient attainable with the electrode is in excess of 0.5 cm s-* (for a typical value of D = 6.5 X 1W cm2s-I), making the electrode attractive for the study of fast reactions under steady-state conditions. Furthermore, there is considerable scope for increasing the available mass-transfer rates by employing
smaller nozzles and electrodes in the MJE and through the use of higher solution flow rates.
ACKNOWLEDGMENT Equipment employed in this work was provided under cumnt grants from SERC (GR/H63739)and the Nuffield Foundation. We thank the University of Warwick Research and Innovation Fund and SERC for support for J.V.M. and Mr. Mark Beeston for many helpful discussions. Received for review January 5, 1994. 1994.’
Accepted March 18,
Abstract published in Aduunce ACS Abstracts. May 1, 1994.
AnaWlChemlsby. Vol. 66, No. 13, Ju& 1, 1994
2119
